deltalake-rb 0.1.2 → 0.1.4

data/README.md

@@ -21,15 +21,15 @@ It can take 5-10 minutes to compile the gem.
 Write data
 
 ```ruby
-df = Polars::DataFrame.new({"a" => [1, 2], "b" => [3.0, 4.0]})
-DeltaLake.write("./data/delta", df)
+df = Polars::DataFrame.new({"id" => [1, 2], "value" => [3.0, 4.0]})
+DeltaLake.write("./events", df)
 ```
 
 Load a table
 
 ```ruby
-dt = DeltaLake::Table.new("./data/delta")
-df2 = dt.to_polars
+dt = DeltaLake::Table.new("./events")
+df = dt.to_polars
 ```
 
 Get a lazy frame
@@ -41,31 +41,31 @@ lf = dt.to_polars(eager: false)
 Append rows
 
 ```ruby
-DeltaLake.write("./data/delta", df, mode: "append")
+DeltaLake.write("./events", df, mode: "append")
 ```
 
 Overwrite a table
 
 ```ruby
-DeltaLake.write("./data/delta", df, mode: "overwrite")
+DeltaLake.write("./events", df, mode: "overwrite")
 ```
 
 Add a constraint
 
 ```ruby
-dt.alter.add_constraint({"a_gt_0" => "a > 0"})
+dt.alter.add_constraint({"id_gt_0" => "id > 0"})
 ```
 
 Drop a constraint
 
 ```ruby
-dt.alter.drop_constraint("a_gt_0")
+dt.alter.drop_constraint("id_gt_0")
 ```
 
 Delete rows
 
 ```ruby
-dt.delete("a > 1")
+dt.delete("id > 1")
 ```
 
 Vacuum
@@ -83,13 +83,13 @@ dt.optimize.compact
 Colocate similar data in the same files
 
 ```ruby
-dt.optimize.z_order(["a"])
+dt.optimize.z_order(["category"])
 ```
 
 Load a previous version of a table
 
 ```ruby
-dt = DeltaLake::Table.new("./data/delta", version: 1)
+dt = DeltaLake::Table.new("./events", version: 1)
 # or
 dt.load_as_version(1)
 ```

data/ext/deltalake/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "deltalake"
-version = "0.1.2"
+version = "0.1.4"
 license = "Apache-2.0"
 authors = ["Andrew Kane <andrew@ankane.org>"]
 edition = "2021"
@@ -14,8 +14,8 @@ crate-type = ["cdylib"]
 arrow = { version = "53", features = ["ffi"] }
 arrow-schema = { version = "53", features = ["serde"] }
 chrono = "0.4"
-delta_kernel = "0.4"
-deltalake = { version = "=0.22.2", features = ["azure", "datafusion", "gcs", "s3"] }
+delta_kernel = "0.6"
+deltalake = { version = "=0.23.0", features = ["azure", "datafusion", "gcs", "s3"] }
 futures = "0.3"
 magnus = "0.7"
 num_cpus = "1"

data/lib/deltalake/table.rb

@@ -188,23 +188,28 @@ module DeltaLake
       JSON.parse(metrics)
     end
 
-    def to_polars(eager: true)
+    def to_polars(eager: true, rechunk: false, columns: nil)
       require "polars-df"
 
       sources = file_uris
-      lf =
-        if sources.empty?
-          Polars::LazyFrame.new
-        else
-          delta_keys = [
-            "AWS_S3_ALLOW_UNSAFE_RENAME",
-            "AWS_S3_LOCKING_PROVIDER",
-            "CONDITIONAL_PUT",
-            "DELTA_DYNAMO_TABLE_NAME"
-          ]
-          storage_options = @storage_options&.reject { |k, _| delta_keys.include?(k.to_s.upcase) }
-          Polars.scan_parquet(sources, storage_options: storage_options)
+      if sources.empty?
+        lf = Polars::LazyFrame.new
+      else
+        delta_keys = [
+          "AWS_S3_ALLOW_UNSAFE_RENAME",
+          "AWS_S3_LOCKING_PROVIDER",
+          "CONDITIONAL_PUT",
+          "DELTA_DYNAMO_TABLE_NAME"
+        ]
+        storage_options = @storage_options&.reject { |k, _| delta_keys.include?(k.to_s.upcase) }
+        lf = Polars.scan_parquet(sources, storage_options: storage_options, rechunk: rechunk)
+
+        if columns
+          # by_name requires polars-df > 0.15.0
+          lf = lf.select(Polars.cs.by_name(*columns))
         end
+      end
+
       eager ? lf.collect : lf
     end
 

data/lib/deltalake/version.rb

@@ -1,3 +1,3 @@
 module DeltaLake
-  VERSION = "0.1.2"
+  VERSION = "0.1.4"
 end

metadata

@@ -1,14 +1,13 @@
 --- !ruby/object:Gem::Specification
 name: deltalake-rb
 version: !ruby/object:Gem::Version
-  version: 0.1.2
+  version: 0.1.4
 platform: ruby
 authors:
 - Andrew Kane
-autorequire:
 bindir: bin
 cert_chain: []
-date: 2024-12-03 00:00:00.000000000 Z
+date: 2025-01-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rb_sys
@@ -24,7 +23,6 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-description:
 email: andrew@ankane.org
 executables: []
 extensions:
@@ -59,7 +57,6 @@ homepage: https://github.com/ankane/delta-ruby
 licenses:
 - Apache-2.0
 metadata: {}
-post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -74,8 +71,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.5.22
-signing_key:
+rubygems_version: 3.6.2
 specification_version: 4
 summary: Delta Lake for Ruby
 test_files: []